Why Pay the Chief? Land Rents and Political Selection in Indonesia
Abstract: Much of modern development efforts are channeled through traditional local governance. Yet, despite their importance as politician-bureaucrats, local leaders are rarely paid a living wage. This paper studies the effect of awarding chiefs cultivation rights over village rice land, a stable revenue generating asset, during their term of office. Using a fuzzy spatial regression discontinuity design, I exploit a historical natural experiment in Java where in the nineteenth century a homogeneous region was split, and in one part chiefs were awarded cultivation rights but not in the other. To measure political outcomes, I collect original data from the field tracing the modern electoral history of 931 chiefs in 193 villages. Higher land rents cause positive chief performance and economic development. Chiefs raise more funds and construct more public goods such that areas under their control are richer and more developed even today. I find evidence consistent with historically positive political selection as a key mechanism. Higher rents attracted better quality chiefs in the past. These chiefs were so effective at educational provision that the entire village today remains more educated. As a result, despite higher land rents attracting a higher quality pool of candidates today, neither candidates nor chiefs today are more selected compared to the average villager. Instead, positive development outcomes today are shaped by the selection of chiefs whose interests are aligned away from supra-village elite interests. Overall, my findings provide evidence that paying local leaders from a stable source of local revenue can be good for economic development.
